Our goals are to develop "chemical genetics", an approach to solving mechanism in cell biology based on use of small molecule tools, and to discover small molecule tools that perturb cell division by novel mechanisms. The small molecules we discover will impact on cancer by revealing target protein/small molecule pairs as the starting point for anti-mitotic drug design. We will develop diversity-oriented synthesis (DOS) pathways that allow synthesis of large libraries of structurally diverse small molecules with complex stereochemistry. Library design will be guided by principles from cheminformatics to maximize diversity. We will screen these libraries for small molecules that perturb cell division, using automated fluorescence microscopy of treated cells, followed by computational analysis of images, to find hits that cause specific phenotypic effects. We will also screen for small molecules that inhibit the function of key proteins known to be involved in cell division using enzymatic assays and assays of protein binding to small molecules immobilized as microarrays. Library synthesis and screening will use a technology platform we developed in the previous funding period. We will optimize the affinity of interesting hits by synthesizing and screening "tuning" libraries that sample chemical space around the original hit. We will start by optimizing a hit we found from a DOS library that targets Eg5, a motor protein required for cell division. We will develop a new method for finding the protein targets of small molecules that cause interesting phenotypic effects, based on in vitro translation of cDNA libraries, and selection of the target protein by its binding to the small molecule immobilized on glass. Having found small molecule tools that perturb cell division by novel mechanisms, we will test their ability to kill cancer cells in vitro.